船舶电力系统容量小，负荷波动性强，船舶电力负荷预测对于船舶电力系统的稳定性和安全性意义重大。本文提出一种能够对船舶电力负荷进行有效且准确的负荷预测方法，在传统的以最小二乘支持向量机作为船舶电力负荷预测方法的基础上，将变种卡方核函数与RBF核函数相结合，同时支持向量机的正则化参数C和标准化参数$ \sigma $的取值对预测精度影响较大，故使用改进的蝴蝶优化算法对预测模型中的参数以及变种卡方核函数的权重系数进行寻优。仿真结果表明，本文提出的预测方法将负荷预测精度提升至97.5119%，因变种卡方核函数的引入，算法能够对特征向量分量权重进行自动调节，并且经蝴蝶优化算法进行参数寻优后的预测模型更为准确，船舶电力负荷预测精度得到进一步提升。
Ship power system has small capacity and strong load fluctuation. Ship power load forecasting is of great significance to the stability and security of ship power system.This paper presents an effective and accurate load forecasting method for ship power load. Based on the traditional least square support vector machine algorithm as the ship power load forecasting method, the variant chi square kernel function and RBF kernel function are combined, and the regularization parameter C and the normalization parameter $ \sigma $ of the support vector machine algorithm have great influence on the prediction accuracy. Therefore, the improved butterfly optimization algorithm is used to optimize the parameters in the prediction model and the weight coefficient of the variant chi square kernel function.The simulation results show that the forecasting method proposed in this paper improves the load forecasting accuracy to 97.5119%. Due to the introduction of the variant chi square kernel function, the algorithm can automatically adjust the weight of the eigenvector component, and the forecasting model optimized by the butterfly optimization algorithm is more accurate, and the forecasting accuracy of the ship power load is further improved.
2023,45(20): 159-166 收稿日期：2022-8-29
DOI：10.3404/j.issn.1672-7649.2023.20.030
分类号：U664.3
基金项目：舟山科技项目（2022C13034）；江苏省产业前瞻与共性关键技术重点项目(BE2018007-2)
作者简介：舒方舟(1998-),男,硕士研究生,研究方向为船舶电力系统
参考文献：
[1] FARAHAT M A, TALAAT M. Short-term load forecasting using curve fitting prediction optimized by genetic algorithms[J]. International Journal of Energy Engineering, 2012, 2(2): 6
[2] 李钷, 李敏, 刘涤尘. 基于改进回归法的电力负荷预测[J]. 电网技术, 2006, 30(1): 99–104
LI P, LI M, LIU D C. Power load forecasting based on improved regression[J]. Power System Technology, 2006, 30(1): 99–104
[3] 魏明奎, 周全, 蔡绍荣, 等. 基于BFGS-FA优化的分数阶灰色模型的中长期负荷预测[J]. 广西大学学报(自然科学版), 2020, 45(2): 270–276
WEI M K, ZHOU Q, CAI S R, et al. Mid-long term load forecasting based on fractional grey model optimized by BFGS-FA[J]. Journal of Guangxi University (Natural Science Edtiton), 2020, 45(2): 270–276
[4] 刘静. 最小二乘SVM在电力推进船舶电力负荷短期预测的应用[J]. 舰船科学技术, 2021, 43(24): 61–63
LIU J. Application of least square SVM in short-term forecasting of electric load of electric propulsion ship[J]. Ship Science and Technology, 2021, 43(24): 61–63
[5] 侯文君, 汪英英, 姚艺新, 等. 支持向量机的船舶短期电力负荷预测方法[J]. 舰船科学技术, 2018(18): 79–81
HOU W j, WANG Y Y, YAO Y X, et al. Short-term load forecasting method based on support vector machine[J]. Ship Science and Technology, 2018(18): 79–81
[6] 李东亮, 张晓锋, 程刚, 等. 多模自适应舰船电力负荷预测算法[J]. 舰船科学技术, 2012, 34(12): 51–54,87
LI D L, ZHANG X F, CHENG G, et al. A multi-mode adaptive load forecasting algorithms of warship[J]. Ship Science and Technology, 2012, 34(12): 51–54,87
[7] 蒋妍. 基于最小二乘支持向量机的聚氯乙烯汽提过程预测模型[J]. 塑料科技, 2020, 48(2): 84–88
JIANG Y. Prediction model of PVC stripping process based on least square support vector machine[J]. Plastics Technology, 2020, 48(2): 84–88
[8] 李腾, 朱希安, 王占刚. 矿井突水水源判别的FOA-LSSVM模型[J]. 北京信息科技大学学报(自然科学版), 2020, 35(3): 41–45.
LI T, ZHU X A, WANG Z G. FOA-LSSVM model for source identification of mine water inrush [ J ]. Journal of Beijing Information Science & Technology University (Natural Science Edition), 2020, 35(3): 41–45.
[9] 王依柔, 张达敏, 徐航, 等. 基于自适应扰动的疯狂蝴蝶算法[J]. 计算机应用研究, 2020, 37(11): 3276–3280
WANG Y R, ZHANG D M, XU H, et al. Crazy butterfly algorithm based on adaptive perturbation[J]. Application Research of Computers, 2020, 37(11): 3276–3280
[10] 姚池. 全电驱船舶电网能量优化管理策略研究[D]. 重庆: 重庆大学, 2017.
